With development of wireless communications technologies and popularization of an intelligent terminal device, big data has emerged. In the big data era, a large amount of spatial-temporal data is generated every day. The spatial-temporal data is data including temporal and spatial information, such as data that is recorded by a base station in a wireless network and that includes a time and a location at which a user makes a phone call, and user sign-in data on a social network. Spatial-temporal data can well reflect a habit and a regularity of work, social activities, and life of a person, and therefore, visualization of spatial-temporal data is a research focus in a visualization field.
A tree map is a data display manner commonly used in the visualization field. FIG. 1 shows a conventional tree map form. Specifically, FIG. 1 is a distribution diagram of population density in the USA. A level-1 rectangle indicates a distribution status of the population in the USA at a state level (for example, California), and a rectangle corresponding to each state is further divided into smaller rectangles indicating a population distribution status of the state at a city level. A conventional tree map can visually and directly display data distribution based on only one dimension (for example, in FIG. 1, the data distribution is represented only based on a region dimension). However, spatial-temporal data includes both time-domain information and region information, and the conventional tree map cannot be applied to visualization of spatial-temporal data.